Scientists have developed a miniaturized microfluidic chip–based technology that they claim could form the basis of a rapid, POC diagnostics platform for quantitatively measuring rare circulating tumor cells (CTCs) or pathogens in clinical samples. The micro-Hall detector (μHD) essentially identifies individual target cells labeled with magnetic nanoparticles (MNPs) as they flow through the channels on a microfluidic chip. The approach provides for high throughput, requires minimal sample processing, and can quantitatively detect individual cells in unprocessed clinical specimens.
The developers at Massachusetts General Hospital and Harvard Medical School say their prototype system could process a throughput of 107 cells/minute, but a few further modifications and changes to the MNP tags could increase throughput to 109 cells/minute, speed up overall assay time, and allow for miniaturization of the setup into a portable platform. Hako Lee, Ph.D., at MGH’s Center for Systems Biology, and colleagues describe the system in Science Translational Medicine, in a paper titled “Ultrasensitive Clinical Enumeration of Rare Cells ex Vivo Using a Micro-Hall Detector.”
The Hall effect relates to the generation of a voltage across an electrical conductor within a magnetic field. The phenomenon is widely used in magnetic field sensors and is a key component of sensing devices such as odometers in cars and the electronic compasses within global positioning systems, the researchers explain.
In order to harness Hall sensing for detecting biomarkers on target cells, they developed a hybrid microfluidic/semiconductor chip structure that uses μHall sensors to pick up the magnetic moment acquired by MNP-labeled cellular targets subjected to a magnetic field. The design comprises an array of eight overlapping μHall sensors that ensure individual cells flowing through the microfluidic channels pass over at least two of the μHall elements. This means each passing cell can be detected regardless of its lateral position in the microfluidic channel, allowing the use of wider channels with less stringent fluidic control than if the cells had to be focused over a single sensor, the researchers note. Effectively, the magnetic moment per cell is directly proportional to both the number of biomarkers labeled on each cell and the magnetic moment of the MNP, but isn’t affected by the size or position of the magnetic objects across the fluidic channel.
Having verified the ability of the system to label and identify a range of different biomarkers on cancer cell lines, the team then moved on to test the μHD platform on whole blood samples that had been spiked with different quantities of cancer cells. The results demonstrated that μHD cancer cell count correlated very well with expected cell numbers, and was effective over a wide range of introduced cell numbers, from 101–105 cancer cells, even though these target cells were vastly outnumbered by the white blood cells (about 106) and red blood cells (about 109) in each 1 mL blood sample.
In contrast with the μHD results, cancer cell detection in whole blood using flow cytometry not only required prior lysis of red blood cells, but showed considerable discrepancies in cell count, the investigators state. “At high cell numbers (>103 cancer cells), flow cytometry errors were caused by false negatives (81% for flow cytometry and 10% for μHD with 106 spiked cells) arising from cell loss during sample preparation and detection steps. At low cell numbers (<103 tumor cells), flow cytometry false positives (900% with 20 spiked cells) were predominantly owing to competing autofluorescence signals from surrounding leukocytes.”
Utility of the platform to identify cells that express more than one specific biomarker was also achieved by combining three different MNPs to screen cultured breast cancer cells for the simultaneous expression of EGFR, Her2, and EpCAM. The results from these analyses, which were confirmed by flow cytometry, showed that the platform could not only detect cells that expressed multiple markers, but also provide an accurate estimate of the level of biomarker expression on each cell.
In two final sets of experiments the team applied the μHD platform to detect CTCs in specimens from patients with advanced ovarian cancer, and also to monitor the effects of anticancer therapy in xenografted mice. First, CTC detection in human ovarian cancer samples using μHD was compared with the clinical gold standard CellSearch platform, which is more sensitive for detecting rare cells than conventional flow cytometry. Each sample was divided in half and one aliquot processed using CellSearch, while the other was magnetically labeled for a panel of four markers, EPCam/Her2, EGFR, and mucin-1 (MUC1), and measured using μHD.
Significantly, CellSearch detected CTCs in only five of 20 ovarian cancer cases, equating to a diagnostic accuracy of 25%. In contrast, the μHD counted a much higher number of CTCs across all patient samples, and detected much higher cell counts in patients with advanced disease who were no longer undergoing therapy or had other aggressive cancer types. “In contrast to CellSearch, the μHD successfully identified CTCs in 100% of patients with evidence of clinical progression (worsening imaging scans or rapidly rising CA-125) as well as stage IV disease, where only 18% of cases were detected with CellSearch,” the investigators state.
To evaluate μHD in monitoring treatment efficacy, the μHD system was used to process fine needle-aspirated tumor samples taken from xenografted mice before and after treatment using either geldanamycin, a heat shock protein 90 (HSP90) inhibitor, or saline (control). The aspirates were labeled with EGFR-specific MNPs and processed for μHD measurement. The resulting data clearly showed a progressive decrease in EGFR expression in human tumor cells taken from the treated mice, but not from the control animals.
“This cost-effective, single-cell analytical technique is well suited to perform molecular and cellular diagnosis of rare cells in the clinic,” the authors conclude. They suggest that while multiplexing is currently limited to the use of 6–7 magnetically defined MNPs, it should be possible to combine Hall sensing with flow cytometry and increase the number of biomarkers that can be tested to over 30.
A number of tweaks to the basic platform will also lead to a number of improvements. These include the use of highly magnetic particles (rather than the weakly magnetic MNPs used in the published study) to label some kinds of biomarkers, and the use of much smaller, scaled down Hall sensors for detecting pathogens in clinical samples. Advances in the acquisition electronics will allow increased cell throughput and dramatically shorten assay time, while integrating the electronics alongside the Hall sensors would both simplify the chip control and aid in miniaturizing the entire system setup for portable operation. “With beneficial features of low-cost electronics and ease of operation, μHDs are poised to offer user-friendly tools for clinicians and investigators seeking rapid insight into disease at the point of care.”